In February 1994, Baxter Healthcare Corporation (BHC) announced the withdrawal of all Immune Globulin Intravenous (IGIV), Gammagard, lots because of reports of hepatitis C transmission. Worldwide, approximately 200 suspected transmissions have been reported since October 1993. We tested 42 Gammagard lots for HCV RNA and assessed the results in relation to the type of anti-HCV screening performed on the starting plasma and the number of case reports received by BHC. HCV RNA was also determined in some US recipients' sera which were kindly provided by the CDC (Centers for Disease Control and Prevention). Gammagard lots prepared between 1988 and November 1993 were assayed for HCV RNA by the reverse transcription-polymerase chain reaction (RT-PCR). The procedure was performed in two independent laboratories with similar nested primers from the 5'-noncoding region of the HCV genome. Of the Gammagard lots produced before and in 1992 from plasma that had not been screened for anti-HCV, laboratory A (Division of Hematology) detected HCV RNA in 1 of 13 lots, whereas laboratory B (Division of Transfusion Transmitted Diseases) found 3 lots positive among 9 tested. Neither laboratory detected HCV RNA in 5 lots produced from plasma mostly or partially screened by EIA-1 (anti-c100-3 test). In contrast, of 24 lots derived mostly from plasma screened by EIA-2 (multiantigen test) and tested by both laboratories, HCV RNA was detected in 16 by laboratory A and in 14 by laboratory B. A total of 20 of the 24 lots examined were positive for HCV RNA by one or both laboratories. This result is highly significant (p<0.0001) when compared to the results obtained by testing lots produced from either unscreened or EIA-1 screened plasma (HCV RNA detected in 3 of 18 lots by at least one laboratory). The Gammagard lot found to have the highest level of HCV RNA by limiting dilution was produced solely from EIA-2 screened plasma and was associated with the most transmissions. Many recipients, who often received multiple RT-PCR positive lots of Gammagard, were found to have HCV RNA in their sera. Thus, screening of plasma by EIA-2 appears to have effected the presence of HCV RNA in Gammagard. Such screening may have removed complexing and/or neutralizing antibodies with the resulting adverse outcome.